The fundamental issue underlying this proposal is the mechanism responsible for associative learning in mammals. Two of the most successful model systems today involve adaptions of the vestibulo-ocular reflex (VOR) and classical conditioning of the nictitating membrane response (NMR). This proposal seeks to significantly extend our understanding of associative learning mechanisms by an in-depth computational and neurophysiological analysis of VOR and its adaptation in the cat combined with an effort to bridge the conceptual gap between this model system and NMR conditioning. Three key questions will be explored: 1. Can a computational model of neural function constrained by known physiology, account for the currently described behavior as well as serve as a heuristic tool in directing further experimentation? This defines a bottom-up approach which has not been utilized on this scale previously in analyzing a specific response system. Our prior modeling work in this project. These predictions include identification of: a) likely anatomical site for learning to take place and b) a physiological substrate for the "teacher" signal in a supervised learning algorithm used to represent reflex plasticity. 2. What is the precise nature of the changes which take place at the single neuron level which ar responsible for the development of motor learning? A two-prong approach to help answer this critical questions is proposed. First, an in-depth characterization of the changes in the spatial characteristic of neurons crucial to expression of the VOR will be conducted. Second, u sing stimulation and lesion techniques the nature of the training signal responsible for guiding plasticity in the reflex will be explore. 3. Can mechanisms accounting for both VOR adaption and classical conditioning of the NMR be unified under a single learning scheme? Different ways in which to map the VOR adaption paradigm to a classical conditioning paradigm will be explored. A clear demonstration of VOR adaptation development under conditions defined by classical conditioning would lead the way to unifying these two significant bodies of investigation into associative learning mechanisms. The benefits of a clear understanding of how the brain express associative learning are enormous form a societal viewpoint; from better recovery of function in stroke patients, to improvements in the behavioral capabilities of the mentally retarded to, potentially, an improvement in the ability of our society, as a whole, to increase its intellectual capacities.